[Mlir-commits] [mlir] [mlir][linalg] Support quantized NCHW/FCHW convolution in img2col (PR #225308)

[Federico Bruzzone]

================
@@ -2754,7 +2754,7 @@ def ConvertConv2DToImg2ColOp : Op<Transform_Dialect,
      ReportTrackingListenerFailuresOpTrait]> {
   let description = [{
     Convert linalg.conv_2d_xxx into linalg.generic (for img2col packing)
-    and linalg.matmul.
+    followed by matrix multiplication or batched matrix-vector multiplication.
----------------
FedericoBruzzone wrote:

I can't follow this documentation in the code. Perhaps it needs updating?

https://github.com/llvm/llvm-project/pull/225308